Abstract The energetic status of the myocardium is compromised in decompensated hypertrophy in the failing heart, with the chemical energy (in the form of the ATP hydrolysis potential) available for the heart to do work diminished compared to normal. The consequences of the observed changes in energetic state on mechanical function are not known. In previous studies we have developed computer models that explain how the depletion of cytoplasmic metabolite pools in the myocardium affects energetic state in heart failure; and a metabolic state-dependent computer model for myocardial mechanics that predicts how these observed changes in energetic status affect mechanical function in vivo. Using these models to interpret data from humans and animal models of cardiac decompensation and heart failure, we predict that metabolic/energetic dysfunction directly causes contractile dysfunction of the myocardium in heart failure. In this project we will test the following hypotheses associated with that prediction: (1.) The primary causes of metabolic/energetic dysfunction in the TAC rat model of heart failure are reduction in mitochondrial capacity for oxidative phosphorylation and pathological depletion of cytoplasmic adenine nucleotides and other key metabolic pools. (2.) Diminished cytosolic ATP and increased inorganic phosphate (associated with impaired energy metabolism) impairs the mechanical function of the heart. (3.) By blocking purine degradation pathways that may be overactive in the chronically stressed and/or periodically ischemic myocardium, we can increase/restore the nucleotide pool and protect the heart against mechanical dysfunction and failure. The three specific aims are built around testing and refining these three hypotheses. Metabolic and functional data from experiments on rat models of hypertrophy and failure will be interpreted based on multi- scale computer models integrating cardiac energetic and mechanics with whole-body cardiovascular function. Hypotheses will be tested and refined based on the ability/inability of the models to simultaneously explain the metabolic and mechanical data from the animal models. This approach expedites the cycle of hypothesis testing (via quantitative comparison of model predictions to experimental observations), hypothesis refinement (redesign and reformulation of models in light of mismatches between predictions and data), and model-guided experimental design. Successful testing of the third hypothesis has the potential to point to whole new classes of pharmacological targets associated with purine nucleotide dephosphorylation, deamination, degradation, and transport.